Microarray-Based Expression Profiling

A DNA microarray measures gene expression by hybridizing labelled nucleic acids from a sample to a dense, ordered grid of complementary probes fixed on a solid surface; the fluorescence at each probe reflects how much of the corresponding transcript is present. Microarrays made genome-scale expression profiling routine in the late 1990s and 2000s and remain a defined, cost-effective platform where the set of transcripts to be measured is known in advance.

Definition

Microarray-based expression profiling quantifies transcript abundance by measuring the hybridization signal of labelled sample nucleic acids to a fixed array of complementary DNA or oligonucleotide probes, yielding a relative expression value for each probed gene.

Scope

This topic covers the design and use of expression microarrays: probe layout, sample labelling and hybridization, image-based signal acquisition, and the normalization and summarization needed to turn raw probe intensities into comparable expression values. It is a methodological reference within transcriptomics and offers no clinical guidance.

Core questions

• How are probes designed and arrayed to represent the genes of interest?
• How does hybridization signal relate to transcript abundance?
• How are raw probe intensities normalized and summarized into per-gene expression values?
• When is a fixed-probe array preferable to direct sequencing of transcripts?

Key concepts

• Probe and array design
• Hybridization and fluorescent labelling
• Two-colour versus single-channel arrays
• Background correction and normalization
• Probe-level summarization (e.g., RMA)
• Cross-hybridization and saturation
• Fixed probe set (closed measurement)

Mechanisms

Labelled complementary DNA or RNA prepared from a sample is applied to an array on which thousands of probes of known sequence are fixed at addressable positions. Complementary molecules hybridize to their probes, and a scanner records the fluorescence intensity at each spot, which is taken as a relative measure of the abundance of the matching transcript. Because intensity is affected by background, probe affinity, and array-to-array variation, raw signals must be background-corrected, normalized across arrays, and summarized from multiple probes per gene into a single expression estimate; the Robust Multi-array Average (RMA) method of Irizarry and colleagues is a widely used summarization framework. Schena and colleagues introduced the cDNA microarray, and Lockhart and colleagues established high-density oligonucleotide arrays as a quantitative expression-monitoring platform.

Clinical relevance

Microarrays produced influential molecular classifications of disease and were the platform behind several early gene-expression signatures used in research and translational settings. As a reference topic this entry explains how array-based expression evidence is generated and normalized; it is not a basis for individual diagnostic or treatment decisions.

Evidence & guidelines

The methodological foundations are the platform papers of Schena and colleagues (cDNA arrays) and Lockhart and colleagues (oligonucleotide arrays), together with normalization and summarization methods such as RMA (Irizarry and colleagues). These are methodological references rather than clinical guidelines.

History

Expression microarrays were introduced in the mid-1990s, with Schena and colleagues demonstrating a cDNA microarray in 1995 and Lockhart and colleagues describing high-density oligonucleotide arrays in 1996. Over the following decade arrays became the dominant tool for genome-scale expression studies, and statistical methods for normalization and probe summarization matured. From the late 2000s, RNA sequencing progressively displaced arrays for discovery work, though arrays persisted where a defined probe set and lower cost were advantageous.

Debates

Microarrays versus RNA sequencing

Arrays measure only the transcripts represented by their fixed probes and have a narrower dynamic range, whereas sequencing measures transcripts directly and can discover novel ones; the trade-off between a defined, cheaper closed assay and an open discovery platform continues to inform platform choice.

Key figures

• Patrick O. Brown
• Mark Schena
• David J. Lockhart
• Rafael Irizarry

Seminal works

• schena-1995
• lockhart-1996
• irizarry-2003

Frequently asked questions

Why can a microarray only measure transcripts it was designed for?

Signal arises from hybridization to probes of known sequence fixed on the array, so any transcript without a matching probe cannot be detected. This makes microarrays a closed assay, in contrast to sequencing, which samples transcripts directly.

What does normalization correct for in microarray data?

It adjusts for technical variation such as differing background, labelling efficiency, and array-to-array intensity differences, so that the resulting expression values reflect biological differences rather than experimental artefacts.